[lammps-users] confused about temperature output in nvt

LAMMPS LAMMPS Mailing List Mirror
1

Hello,

I am trying to run nvt, and from reading the users manual, it states:

fix ID group-ID style_name keyword value
 *temp* values = Tstart Tstop Tdamp
    Tstart,Tstop = external temperature at start/end of run
    Tdamp = temperature damping parameter (time units)

so if I say  fix 1 all nvt temp 300.0 1400.0 150.0 the simulation should run from 300 - 1400 K. 
however, with the following input I get very strange temperature output results. What is going on?

Input file:

1 units metal
2 boundary p p p
3
4 atom_style atomic
5
6 lattice custom 3.015 a1 1.0 0.0 0.0 a2 0.0 1.0 0.0 a3 0.0 0.0 1.0 &
7 basis 0.0 0.0 0.0 basis 0.5 0.5 0.5 origin 0 0 0
8
9 region mysphere sphere 0.0 0.0 0.0 4.0 side in
10 #region mybox block 0.0 20.0 0.0 20.0 0.0 20.0 side in
11 #region bounds union 2 mysphere mybox
12
13 create_box 2 mysphere
14 create_atoms 2 region mysphere basis 1 1
15 mass 1 1.0
16 mass 2 1.0
17
18
19 velocity all create 0.0 87287 rot yes dist gaussian
20
21 pair_style eam/fs
22 pair_coeff * * /home/ualacc/lammps/potentials/NiTi.eam.fs Ni Ti
23
24 pair_style eam/fs
25 pair_coeff * * /home/ualacc/lammps/potentials/NiTi.eam.fs Ni Ti
26
27 neighbor 0.3 bin
28 neigh_modify delay 0 every 1 check yes
29
30 compute 1 all temp
31 fix 1 all nvt temp 300.0 1400.0 150.0
32 thermo 10
33 run 100
34
35 neigh_modify delay 0 every 1 check yes
36
37 dump 1 all atom 100 dump.min

Output:

32 Step Temp E_pair E_mol TotEng Press
33 0 15975.788 4453.0924 0 5553.7548 707183.84
34 10 14127.444 4612.6843 0 5586.0038 715033.54
35 20 12851.069 4702.3746 0 5587.7575 714098.95
36 30 16619.798 4441.7719 0 5586.8039 701947.68
37 40 20525.376 4169.722 0 5583.8313 687086.87
38 50 23764.589 3939.3792 0 5576.6562 671108.04
39 60 26769.298 3723.5124 0 5567.8008 656791.96
40 70 30209.449 3480.3548 0 5561.6546 639669.11
41 80 32499.326 3317.0985 0 5556.1609 628316.83
42 90 35049.26 3138.7191 0 5553.4609 612033.14
43 100 36729.717 3019.0626 0 5549.5806 600861.75
44 Loop time of 0.0707879 on 1 procs for 100 steps with 534 atoms

Why are the temperature values so unreasonable? I don’t understand.

Thanks in advance,

2

hello amanda,

Hello,

I am trying to run nvt, and from reading the users manual, it states:

fix ID group-ID style_name keyword value
temp values = Tstart Tstop Tdamp
    Tstart,Tstop = external temperature at start/end of run
    Tdamp = temperature damping parameter (time units)

so if I say fix 1 all nvt temp 300.0 1400.0 150.0 the simulation should run
from 300 - 1400 K.
however, with the following input I get very strange temperature output
results. What is going on?

before trying a temperature ramp, you first have to
equilibrate your system to the initial temperature of 300K.
very obviously, your input has high potential energy and
pressure, so perhaps there is something else wrong.

you should investiage that first and then try the temperature ramp.

axel.

3

fix all nvt temp 300.0 1400.0 150.0
run 100

It's the 150 that is bad. You're asking
the thermostat to change the temp from 300 to 1400
in 100 timesteps, yet setting the damping constant
to 150 = 150,000 timesteps in metal units.
The thermostat won't work with those settings.
See the doc page for details.

More generally, it makes little sense to ask any
thermostat to change the temperature by 1100 degrees
in 100 timesteps. It's too drastic.

Steve